Temperature Stabilised Bearings - how is the stabilising done ?

[David Jupp]

I've just been advised that temperature stabilised bearings will no longer be supplied in particular sizes due to 'lack of demand'.  Manufacturer takes the view that 'standard' bearings ought to be OK for almost all applications, and for others it is just tough.

The application I'm thinking of uses S2 bearing in the cooler part of the system and S4 in the hotter part.  These bearings are moderately large and expensive, so stripping and re-working standard bearings may not be as daft as it sounds. 

Does anyone have any detail of how rolling bearings are heat stabilised?  Can anyone point me to a resource on this matter?

My other thought is to work with a specialist company to move to ceramic bearings.

A secondary question - the temperature stabilisation is claimed to be about dimensional stability, BUT what exactly does this mean?  What happens if the bearing gets too hot (apart from thermal expansion)?       I ask because the application in question is surrounded by 'myth and magic', but nobody seesm to really know on what basis the original design decisions were made nor if they are critical.

Any input is valued!

[NeoTech]

Can only relate to what we do when making custom knifes during heat treat.. Then a temperature stabilizing process is done by packing the hot item in Co2 ice. Its also called a Subzero heat treat or zero stabilisation..

It makes crap really frickin hard.. especially 440C steels. And when the steel is heated backup again it really doesnt loose any tolerances. Unless its closing in on temperatures that would omit the heat treat completely. (800C and above in my case with knifes).

[David Jupp]

Thanks - re-heat treating a finish ground part isn't really very practical.  I've since found some information that suggests that different steels may be used for the higher S grade bearing races (not sure about the balls).  Either way it isn't looking very promising to 'upgrade' an existing bearing, it seems the temperature stabilisation happens much too early in manufacturing to be able to do it as an after thought.

Looks like the options are find an alternative manufacturer, or move to ceramic/hybrid bearings.

[Lew_Merrick_PE]

David,

I am not sure about the process of which you are speaking.  I know of bearing elements designed and made for operation at a specific temperature, but I have never heard that referred to as stabilization.  I am more familiar with the term as applied to plastics and aluminum materials where you "soak" them at a somewhat elevated temperature as part of a stress relieving process.

Most steels have a Coefficient of Thermal Expansion (CTE) that runs in the close neighborhood of .000006 in/in/°F.  Thus, at an operating temperature of 570°F, they would expand (500 X .000006 =) .003 in/in of dimensional length (glossing over some geometric relationships for the sake of simplicity).  If the (say) inner bearing race had a neutral axis diameter of 3.250 inches (not uncommon in 3.000 ID bearings), then the race would expand to (pi X 3.250 + 3.250 X pi X .003 =) 10.2408 inches in length making for a (10.2408/pi =) 3.2598 neutral axis diameter -- which could really screw up the fit of your bearing elements.

It has been many years since I have done this, but bearings are designed for such operating conditions.  There is a lot of art in such design work.  But I have never heard this referred to as heat stabiliziation...

[David Jupp]

Hi Lew - that is what the bearing suppliers we have talked to seem to call it, though there may be other names for it.  The one thing that should be universal is the S number (regardless of what you call how it is achieved).  Ratings up to S5 exist, but not necessarily for all bearing types/sizes.

S0 for operating temperature up to 302ºF
S1 for operating temperature up to 392ºF
S2 for operating temperature up to 482ºF
S3 for operating temperature up to 572ºF
S4 for operating temperature up to 662ºF
S5 for operating temperature up to 752ºF

The clearance class is important too and is specified as C3. In this application the temperature comes from the environment not really from any localised heating of the bearing elements, so thermal expansion shoudln't be too much of an issue.

The bearings used are double row shperical roller bearings of moderate size - used to be able to get S2 and S4 without problem, but other applications which needed this have apprently disappeared or have been dealt with differently.  Our application is now left 'out on a limb'....

[Lew_Merrick_PE]

Hi David -- I suspect that we are, once again, people separated by a common tongue.  Is it a wrench or is it a spanner, etc?

If you notice, your classes of bearings have  90°F operating spans.  I suspect that (say) an S5 bearing would rattle like a castanet at room temperature.  There is almost certainly a mechanical/thermal preload allowance in each class such that the shrink-fit assembly will take up some (if not most) of the CTE/thermal expansion load in a manner similar to (mechanical) pre-loading of a bolt which reduces the rattle at lower temperatures -- at least that is how I remember it.

Most of the things I have done requiring thermally specific bearings in recent years have gone on high-altitude aircraft and spacecraft which have opposite direction in their operating temperature requirements.  There the issue is less CTE/thermal contraction than it is low-temperature embrittlement of steels.

I suspect (but do not know directly) that you could approach it from the other side of the interface by adjusting your shaft and bore sizes to create the correct preload.  That would certainly be simpler than diassembling bearings and making new race components.  ???

[David Jupp]

Lew, thanks for your thoughts.  I'm talking specifically about temperature stabilisation NOT thermal expansion issues.  This is to do with the steels retaining hardness and hence not distorting at the operating temperature.  I.E. it's a heat treatment and/or steel composition issue (and maybe some other 'magic').

The S value and the C value (these bearings have both specified) relate to different things.  CTE shoudln't really be a problem as steel balls in steel cages in a uniform temperature environment (albeit hot) should not alter clearance very much from that at room temp - the brass cage may be affected of course.

Am trying both alternate supplier of conventional bearings and a move to hybrid (steel race ceramic rollers, rated for 450 C - 842 F).

It actually turns out that S3 would probably be adequate for our application - the S4 spec comes from original design which envisaged operation at up to 350 C, in reality it is under 300 C.

[Lew_Merrick_PE]

David,

1) Just the fact that you have steel races and bearing elements all with the same CTE does not make the delta-T issue irreverent.  The sizes of each element is different, so the expansion of each element is different.  At least that was the primary issue back when I was building and designing (in the evolution of my career) rotary elements in hot gas-flow environments.

2) Think about the heat treating process.  First there is hardening where you heat the metal up to some soak temperature, increase the temperature to a second soak temperature (usually just above the Curie Point), and then quench it to force the crystal structure of the metal to remain in the (high temperature modified) form.  This is usually followed in modern commercial practice by a quench & soak at some seriously sub-zero (cryogenic) operation that, as I was taught in the 70's, converts more of the crystal structure into the high temperature modified form before being returned to room temperature.  We now have a part with high hardness, high (stress-based) strength, and high brittleness because of the internal strains (i.e. stresses) inherent in the unbalanced high temperature modified crystal structure.  Those stresses (strains) are relieved by tempering.  The higher the soak temperature used in tempering, the more of the stresses are relieved and the hardness, strength, and brittleness are reduced.  (And, in modern practice, tempering often includes more cryogenic processing.)  The variations in temperatures and time of soak vary by the chemical composition of the alloy.

High-speed steels are a class of alloys that retain their hardness, strength, and reduced brittleness (toughness) to higher operating temperatures than traditional carbon steels.  Other classes of alloys (hot-work, tungsten alloy, etc.) have been developed for similar reasons.  Thus, if you need a part to retain (say) Rc-62/60 hardness at some elevated temperature, it is more a matter of selecting the appropriate alloy (with the concomitant heat treating regimen) than applying a process to an alloy without the ability to maintain physical properties at the operating temperature.  It would not surprise me (not being an expert in this field) that there is some secondary thermal processing required to establish properties at a given elevated operating temperature, but I have never heard of such a thing being called stabilization.

At the risk of sounding like a hopeless pedant, I looked up heat stabilization in my 1984 Edition of the (26 volume) ASM Metals Handbook and found nothing appropriate to bearings.  (Most of what I found deals with alloy composition for stability at various operating temperatures.)  I have the standard machinist/tool & die/blacksmith background in metallurgy of my apprenticeship coupled with 2+ years of having a formally trained and well-experienced METALLURGIST (who led metallurgical R&D for the US Army for more than two decades) as my adviser in college.  I have held my own with metallurgists over the years in a wide array of projects -- but I am not a metallurgist -- only a general-purpose mechanical design engineer.

[David Jupp]

Lew,

I have a degree in Metallurgy (for whatever that might be worth) - however I was not at all clear on what 'temperature stabilised' actually MEANT (that's why I posted here).  It is a term usd by the bearing manufacturers, but they don't to my mind explain it at all clearly.  Like you, the term did not match with my education in these matters.

I take your point about different scale of rolling elements vs races.   I'll consider further.

I have subsequently found a little bit more information from manufacturers, but not a lot.  It seems likely that in this context the term refers to a mixture of material selection and heat treatment.

[Lew_Merrick_PE]

David,

One of the companies that was a bread & butter client back in the 80's was The Rocket Research Company.  They had bumper stickers printed up that read: But what if you are a rocket scientist and you still can't figure it out?  Unfortunately, I was not working on site for them when they handed them out...